Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Michael Fuhr
On Wed, Jul 27, 2005 at 12:56:15AM -0400, Tom Lane wrote: Michael Fuhr [EMAIL PROTECTED] writes: Could this be platform-specific? Seems that way. I tried it on HPUX 10.20/HPPA/gcc 2.95.3. No luck on FreeBSD 4.11-STABLE/i386/gcc 2.95.4. The box that does have a problem is Solaris

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Neil Conway
Michael Fuhr wrote: Can anybody else reproduce the problem? I couldn't repro it, on x86 / Debian unstable / gcc 3.4.4, with current CVS sources. -Neil ---(end of broadcast)--- TIP 2: Don't 'kill -9' the postmaster

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Kevin McArthur
The target system for my reproduction is on FreeBSD. Though I sitll cannot get the initdb started one to work the first time around. I cannot reproduce the bug on my linux machine, however I cannot test the full initdb procedure until tomorrow on freebsd. This error has come up in the last

[HACKERS] Idea - optimising (left) joins?

2005-07-27 Thread Dawid Kuroczko
Hello. I was just wondering, assume we have such tables: CREATE TABLE data ( foo text, somename_id integer not null references (somenames) ); CREATE TABLE somenames ( somename_id serial PRIMARY KEY somename text NOT NULL ); And a view: CREATE someview AS SELECT foo,somename FROM

Re: [HACKERS] regression failure on latest CVS

2005-07-27 Thread ohp
In the meantime would'nt it be nice to try to understand what happens and correct it? I'm a bit afraid that 8.1 is not used on unixware because people don't have/want the patch installed. Regards On Tue, 26 Jul 2005, Larry Rosenman wrote: Date: 26 Jul 2005 08:52:01 -0500 From: Larry Rosenman

Re: [HACKERS] Idea - optimising (left) joins?

2005-07-27 Thread Tino Wildenhain
Am Mittwoch, den 27.07.2005, 10:40 +0200 schrieb Dawid Kuroczko: Hello. I was just wondering, assume we have such tables: CREATE TABLE data ( foo text, somename_id integer not null references (somenames) ); CREATE TABLE somenames ( somename_id serial PRIMARY KEY somename

Re: [HACKERS] regression failure on latest CVS

2005-07-27 Thread Larry Rosenman
ohp@pyrenet.fr wrote: In the meantime would'nt it be nice to try to understand what happens and correct it? I'm a bit afraid that 8.1 is not used on unixware because people don't have/want the patch installed. Regards Let's see what they find, first. They may have a work-around (they

Re: [HACKERS] VACUUM DATABASE

2005-07-27 Thread Robert Treat
On Tuesday 26 July 2005 16:53, Alvaro Herrera wrote: On Tue, Jul 26, 2005 at 09:30:20PM +0100, Simon Riggs wrote: I'd like to suggest altering the syntax of VACUUM so that it is possible to issue the command VACUUM DATABASE. The keyword DATABASE would be optional, to allow backward

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Michael Fuhr
On Wed, Jul 27, 2005 at 12:19:51AM -0700, Kevin McArthur wrote: The target system for my reproduction is on FreeBSD. What version of FreeBSD? What compiler and version? So far I haven't been able to reproduce the problem on FreeBSD 4.11-STABLE/ i386/gcc 2.95.4. Though I sitll cannot get the

Re: [HACKERS] regression failure on latest CVS

2005-07-27 Thread Andrew Dunstan
ohp@pyrenet.fr wrote: In the meantime would'nt it be nice to try to understand what happens and correct it? I'm a bit afraid that 8.1 is not used on unixware because people don't have/want the patch installed. All the evidence is that this is a compiler bug. The apparent workaround

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Michael Fuhr
On Wed, Jul 27, 2005 at 12:19:51AM -0700, Kevin McArthur wrote: This error has come up in the last week or so, and my suspicion remains that its caused by something to do with roles but that could be way wrong. The FreeBSD machines were confirmed to work as of about a week ago ( i

Re: [HACKERS] regression failure on latest CVS

2005-07-27 Thread Larry Rosenman
On Jul 27 2005, Andrew Dunstan wrote: ohp@pyrenet.fr wrote: In the meantime would'nt it be nice to try to understand what happens and correct it? I'm a bit afraid that 8.1 is not used on unixware because people don't have/want the patch installed. All the evidence is that this is a

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Kevin McArthur
I cannot repoduce your experience with this bug. No matter what I do, reconnect session or otherwise, it never returns a proper oid on the newer cvs vers (I suspect it may be related to the roles update) Kevin - Original Message - From: Michael Fuhr [EMAIL PROTECTED] To: Kevin McArthur

Re: [HACKERS] regression failure on latest CVS

2005-07-27 Thread Larry Rosenman
On Jul 27 2005, Bruce Momjian wrote: Andrew Dunstan wrote: ohp@pyrenet.fr wrote: In the meantime would'nt it be nice to try to understand what happens and correct it? I'm a bit afraid that 8.1 is not used on unixware because people don't have/want the patch installed.

Re: [HACKERS] More buildfarm stuff

2005-07-27 Thread Palle Girgensohn
--On tisdag, juli 26, 2005 15.17.57 -0400 Tom Lane [EMAIL PROTECTED] wrote: Larry Rosenman ler@lerctr.org writes: On Jul 26 2005, Jim C. Nasby wrote: So the question now is: how do we fix the issue with threaded python? how do we get libc_r into the mix on FreeBSD 4.11? A possible

Re: [HACKERS] regression failure on latest CVS

2005-07-27 Thread Bruce Momjian
Andrew Dunstan wrote: ohp@pyrenet.fr wrote: In the meantime would'nt it be nice to try to understand what happens and correct it? I'm a bit afraid that 8.1 is not used on unixware because people don't have/want the patch installed. All the evidence is that this is a

Re: [HACKERS] Idea - optimising (left) joins?

2005-07-27 Thread Stephan Szabo
On Wed, 27 Jul 2005, Dawid Kuroczko wrote: Hello. I was just wondering, assume we have such tables: CREATE TABLE data ( foo text, somename_id integer not null references (somenames) ); CREATE TABLE somenames ( somename_id serial PRIMARY KEY somename text NOT NULL ); And

Re: [HACKERS] Interesting COPY edge case...

2005-07-27 Thread Bruno Wolff III
On Tue, Jul 26, 2005 at 17:56:42 -0400, Chris Browne [EMAIL PROTECTED] wrote: There is evidently Something Strange about the state of stdout when it is referenced inside a stored procedure. I suspect this is related to trusted PLs not being able to write files. It does seem like a problem

Re: [HACKERS] ENUM type

2005-07-27 Thread Andrew Dunstan
Tom Lane wrote: Jim C. Nasby [EMAIL PROTECTED] writes: ... what I was wondering about is creating a 'type' that is a rollup for: - create parent table with int id field and text and indexes - add RI to base table - add triggers/views/rules/other glue to make the id field hidden

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Andrew - Supernews
On 2005-07-27, Michael Fuhr [EMAIL PROTECTED] wrote: So far the problem does seem to be specific to whatever PL/pgSQL's is doing, and it affects ROW_COUNT as well as RESULT_OID. I haven't been able to reproduce the problem with PL/Tcl or with C and SPI. src/pl/plpgsql/src/pl_exec.c, function

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Andrew Dunstan
Andrew - Supernews wrote: On 2005-07-27, Michael Fuhr [EMAIL PROTECTED] wrote: So far the problem does seem to be specific to whatever PL/pgSQL's is doing, and it affects ROW_COUNT as well as RESULT_OID. I haven't been able to reproduce the problem with PL/Tcl or with C and SPI.

[HACKERS] Integrated autovacuum

2005-07-27 Thread Joshua D. Drake
Hello, Just for clarification, will the new integrated autovacuum require that statistics are on? Sincerely, Joshua D. Drake -- Your PostgreSQL solutions company - Command Prompt, Inc. 1.800.492.2240 PostgreSQL Replication, Consulting, Custom Programming, 24x7 support Managed Services,

Re: [HACKERS] Integrated autovacuum

2005-07-27 Thread Alvaro Herrera
On Wed, Jul 27, 2005 at 12:53:40PM -0700, Joshua D. Drake wrote: Just for clarification, will the new integrated autovacuum require that statistics are on? Yes. Row-level stats too. -- Alvaro Herrera (alvherre[a]alvh.no-ip.org) I'm always right, but sometimes I'm more right than other

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Kevin McArthur
Changing just the one appears to resolve the oid bug. Should probably talk to neilc to see why he changed it. I will pass along a patch for this particular case to -patches shortly Kevin McArthur - Original Message - From: Andrew Dunstan [EMAIL PROTECTED] To: [EMAIL PROTECTED] Cc:

[HACKERS] wal_buffer tests in

2005-07-27 Thread Josh Berkus
Folks, I ran a wal_buffer test series. It appears that increasing the wal_buffers is indeed very important for OLTP applications, potentially resulting in as much as a 15% average increase in transaction processing. What's interesting is that this is not just true for 8.1, it's true for

Re: [HACKERS] --enable-thread-safety on Win32

2005-07-27 Thread Dave Page
Did anyone get a chance to think about this? I'd like to fix this for 8.1, but it should also make life easy with the new libpq based ODBC driver improvements if I can produce an appropriate patch sooner rather than later! Regards, Dave. -Original Message- From: [EMAIL PROTECTED]

Re: [HACKERS] Integrated autovacuum

2005-07-27 Thread Joshua D. Drake
Alvaro Herrera wrote: On Wed, Jul 27, 2005 at 12:53:40PM -0700, Joshua D. Drake wrote: Just for clarification, will the new integrated autovacuum require that statistics are on? Yes. Row-level stats too. Great Thanks... Could I get a better explanation of the following:

Re: [HACKERS] Integrated autovacuum

2005-07-27 Thread Martín Marqués
El Mié 27 Jul 2005 17:24, Alvaro Herrera escribió: On Wed, Jul 27, 2005 at 12:53:40PM -0700, Joshua D. Drake wrote: Just for clarification, will the new integrated autovacuum require that statistics are on? Yes. Row-level stats too. Will there be a way to ballance the amount of stats

Re: [HACKERS] Integrated autovacuum

2005-07-27 Thread Alvaro Herrera
On Wed, Jul 27, 2005 at 06:05:26PM -0300, Martín Marqués wrote: El Mié 27 Jul 2005 17:24, Alvaro Herrera escribió: On Wed, Jul 27, 2005 at 12:53:40PM -0700, Joshua D. Drake wrote: Just for clarification, will the new integrated autovacuum require that statistics are on? Yes.

Re: [HACKERS] Integrated autovacuum

2005-07-27 Thread Alvaro Herrera
On Wed, Jul 27, 2005 at 02:07:28PM -0700, Joshua D. Drake wrote: Great Thanks... Could I get a better explanation of the following: #autovacuum_vacuum_scale_factor = 0.4 # fraction of rel size before vacuum #autovacuum_analyze_scale_factor = 0.2 # fraction of rel size before analyze

Re: [HACKERS] wal_buffer tests in

2005-07-27 Thread Mark Wong
On Wed, 27 Jul 2005 13:30:01 -0700 Josh Berkus josh@agliodbs.com wrote: Folks, I ran a wal_buffer test series. It appears that increasing the wal_buffers is indeed very important for OLTP applications, potentially resulting in as much as a 15% average increase in transaction processing.

Re: [HACKERS] Integrated autovacuum

2005-07-27 Thread Martín Marqués
El Mié 27 Jul 2005 18:23, Alvaro Herrera escribió: On Wed, Jul 27, 2005 at 06:05:26PM -0300, Martín Marqués wrote: Will there be a way to ballance the amount of stats the autovacuum gets? Something like the analyze parameters that the contrib version has, but integrated in

Re: [HACKERS] wal_buffer tests in

2005-07-27 Thread Simon Riggs
On Wed, 2005-07-27 at 13:30 -0700, Josh Berkus wrote: I ran a wal_buffer test series. It appears that increasing the wal_buffers is indeed very important for OLTP applications, potentially resulting in as much as a 15% average increase in transaction processing. What's interesting is

[HACKERS] Some new list.c primitives

2005-07-27 Thread Tom Lane
Neil (or anyone else with an opinion), I'm finding several uses in the planner for some new List primitives defined as below. I'd like to push these into list.c, but before that, has anyone got any serious objections? How about suggestions for better names? regards, tom

Re: [HACKERS] Integrated autovacuum

2005-07-27 Thread Alvaro Herrera
On Wed, Jul 27, 2005 at 06:35:31PM -0300, Martín Marqués wrote: El Mié 27 Jul 2005 18:23, Alvaro Herrera escribió: On Wed, Jul 27, 2005 at 06:05:26PM -0300, Martín Marqués wrote: Will there be a way to ballance the amount of stats the autovacuum gets? Something like the analyze

Re: [HACKERS] Checkpoint cost, looks like it is WAL/CRC

2005-07-27 Thread Simon Riggs
On Tue, 2005-07-26 at 19:15 -0400, Tom Lane wrote: Josh Berkus josh@agliodbs.com writes: We should run tests with much higher wal_buffers numbers to nullify the effect described above and reduce contention. That way we will move towards the log disk speed being the limiting factor, patch or

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Michael Fuhr
On Wed, Jul 27, 2005 at 01:20:29PM -0700, Kevin McArthur wrote: Changing just the one appears to resolve the oid bug. Should probably talk to neilc to see why he changed it. Initializing isnull to false in exec_stmt_getdiag() appears to fix the bug on my Solaris 9 box as well. I'd guess the

Re: [HACKERS] Checkpoint cost, looks like it is WAL/CRC

2005-07-27 Thread Josh Berkus
Simon, I guess I'd be concerned that the poor bgwriter can't do all of this work. I was thinking about a separate log writer, so we could have both bgwriter and logwriter active simultaneously on I/O. It has taken a while to get bgwriter to perform its duties efficiently, so I'd rather not

Re: [HACKERS] Integrated autovacuum

2005-07-27 Thread Josh Berkus
Alvaro, Also, somebody (Rod Taylor I think) proposed changed the variable names to vacuum_auto_vacuum_scale_factor vacuum_auto_analyze_scale_factor I see what Rod's getting at, but I find that version of the option less readable ... -- --Josh Josh Berkus Aglio Database Solutions San

Re: [HACKERS] VACUUM DATABASE

2005-07-27 Thread Simon Riggs
On Wed, 2005-07-27 at 00:07 -0400, Robert Treat wrote: On Tuesday 26 July 2005 16:53, Alvaro Herrera wrote: On Tue, Jul 26, 2005 at 09:30:20PM +0100, Simon Riggs wrote: I'd like to suggest altering the syntax of VACUUM so that it is possible to issue the command VACUUM DATABASE. The

[HACKERS] Sanity Check?

2005-07-27 Thread Larry Rosenman
SCO is seeing the following failure without -O, but no failure with -O: The sanity_check diffs show: *** ./expected/sanity_check.out Wed Jul 27 17:58:12 2005 --- ./results/sanity_check.out Wed Jul 27 18:09:41 2005 *** *** 17,22 --- 17,24 circle_tbl | t

Re: [HACKERS] Sanity Check?

2005-07-27 Thread Tom Lane
Larry Rosenman ler@lerctr.org writes: SCO is seeing the following failure without -O, but no failure with -O: *** ./expected/sanity_check.out Wed Jul 27 17:58:12 2005 --- ./results/sanity_check.out Wed Jul 27 18:09:41 2005 *** *** 17,22 --- 17,24 circle_tbl

Re: [HACKERS] RESULT_OID Bug

2005-07-27 Thread Kevin McArthur
Sounds plausible. Can anyone comment on whether the other instances of isnull=false being removed need to be reset to =false Kevin McArthur - Original Message - From: Michael Fuhr [EMAIL PROTECTED] To: Kevin McArthur [EMAIL PROTECTED] Cc: Andrew Dunstan [EMAIL PROTECTED]; [EMAIL

Re: [HACKERS] VACUUM DATABASE

2005-07-27 Thread Alvaro Herrera
On Wed, Jul 27, 2005 at 11:41:24PM +0100, Simon Riggs wrote: Forgive me if this is wrong, but I took that Alvaro was applying a reductio ad absurdum argument (i.e. taking the piss). I laughed heartily at the thought of LAZY becoming a PostgreSQL keyword. Ah, well, actually you are wrong, I

Re: [HACKERS] Sanity Check?

2005-07-27 Thread Alvaro Herrera
Why does the sanity check test start with a VACUUM? Why not a VACUUM FULL? -- Alvaro Herrera (alvherre[a]alvh.no-ip.org) Los románticos son seres que mueren de deseos de vida ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ?

Re: [HACKERS] ORDER BY field not in return list

2005-07-27 Thread Marc G. Fournier
On Mon, 25 Jul 2005, Jeffrey W. Baker wrote: On Mon, 2005-07-25 at 19:08 -0300, Marc G. Fournier wrote: On Mon, 25 Jul 2005, Jeffrey W. Baker wrote: On Mon, 2005-07-25 at 18:11 -0300, Marc G. Fournier wrote: Just curious as to whether or not a warning or something should be issued in a

Re: [HACKERS] Some new list.c primitives

2005-07-27 Thread Gavin Sherry
On Wed, 27 Jul 2005, Tom Lane wrote: Neil (or anyone else with an opinion), I'm finding several uses in the planner for some new List primitives defined as below. I'd like to push these into list.c, but before that, has anyone got any serious objections? How about suggestions for better

Re: [HACKERS] ORDER BY field not in return list

2005-07-27 Thread Josh Berkus
Marc, Basically, from how everyone has explained it, the ORDER BY will be done after all the JOINs are done, and the product of the joins are complete ... for it to be performed on a field not in the SELECT field clause, then those fields have to be loaded into memory, *then* ORDERed, and

Re: [HACKERS] Some new list.c primitives

2005-07-27 Thread David Fetter
On Wed, Jul 27, 2005 at 06:01:21PM -0400, Tom Lane wrote: Neil (or anyone else with an opinion), I'm finding several uses in the planner for some new List primitives defined as below. I'd like to push these into list.c, but before that, has anyone got any serious objections? How about

[HACKERS] Autovacuum to-do list

2005-07-27 Thread Alvaro Herrera
Hackers, This is a list of things people mentioned as interesting to do for vacuum/autovacuum, during the last autovacuum discussion thread. Some of them are wishful thinking, others are doable. Neil, Gavin: both of you mentioned that you didn't like autovacuum as a long term solution. May I

Re: [HACKERS] Some new list.c primitives

2005-07-27 Thread Tom Lane
David Fetter [EMAIL PROTECTED] writes: How about list_push for both of these? list_push to me would connote the functionality of lappend, ie, unconditionally add the item to the list. regards, tom lane ---(end of

Re: [HACKERS] --enable-thread-safety on Win32

2005-07-27 Thread Bruce Momjian
Dave Page wrote: Did anyone get a chance to think about this? I'd like to fix this for 8.1, but it should also make life easy with the new libpq based ODBC driver improvements if I can produce an appropriate patch sooner rather than later! I have thought about it and it needs to be addressed

Re: [HACKERS] Some new list.c primitives

2005-07-27 Thread Neil Conway
Gavin Sherry wrote: list_add() doesn't really describe what it does. I agree -- the functionality itself is fine, of course, but it would be nice to have a better name. I was thinking either list_cond_add() or list_merge(). What about list_append_distinct()? (And

Re: [HACKERS] Autovacuum to-do list

2005-07-27 Thread Gavin Sherry
On Wed, 27 Jul 2005, Alvaro Herrera wrote: Hackers, This is a list of things people mentioned as interesting to do for vacuum/autovacuum, during the last autovacuum discussion thread. Some of them are wishful thinking, others are doable. Neil, Gavin: both of you mentioned that you didn't

Re: [HACKERS] [Testperf-general] dbt2 opteron performance

2005-07-27 Thread Josh Berkus
Mark, I'm starting to get results with dbt2 on a 4-way opteron system and wanted to share what I've got so far since people have told me in the past that this architecture is more interesting than the itanium2 that I've been using. This 4-way has 8GB of memory and four Adaptec 2200s

Re: [HACKERS] Some new list.c primitives

2005-07-27 Thread Tom Lane
Neil Conway [EMAIL PROTECTED] writes: I agree -- the functionality itself is fine, of course, but it would be nice to have a better name. Those were just the first names that came to mind, and of course the reason I asked is that I felt they could be improved upon... I was thinking either

Re: [HACKERS] Autovacuum to-do list

2005-07-27 Thread Alvaro Herrera
On Thu, Jul 28, 2005 at 11:42:22AM +1000, Gavin Sherry wrote: On Wed, 27 Jul 2005, Alvaro Herrera wrote: * Change how we track the need to vacuum I don't really think the use of row level stats are the best indicator for a need to vacuum. I wonder if we could keep a

Re: [HACKERS] [Testperf-general] dbt2 opteron performance

2005-07-27 Thread Mark Wong
On Wed, Jul 27, 2005 at 07:32:34PM -0700, Josh Berkus wrote: Mark, I'm starting to get results with dbt2 on a 4-way opteron system and wanted to share what I've got so far since people have told me in the past that this architecture is more interesting than the itanium2 that I've been

[HACKERS] Possible planner deficiency?

2005-07-27 Thread kris . shannon
I have a situation where i need to select a couple of rows from an inherited table collection. The statement I'm using is: SELECT * FROM parent NATURAL JOIN interesting where interesting is a 1 column temporary table with the primary key's of the rows I'm interested in. All the child tables

Re: [HACKERS] Some new list.c primitives

2005-07-27 Thread Neil Conway
Tom Lane wrote: How about list_append_distinct and list_concat_distinct? Those names are fine with me. -Neil ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq